Keep the misses
Before trying the practice, write what you expect to happen and when you will count it. Afterward, write what actually happened. Keep these records separate. A remembered expectation is unusually easy to repair after the fact.
Make three columns if you can: expected, observed, other explanations. A change in routine, a friend's reminder, a quieter room, more sleep, a deadline—these are not disappointments. They are candidate causes.
Count a miss as a miss. If a cue did not lead to the action, do not enlarge the meaning of success until it does. You can revise the next trial without editing the last one.
Try the same small behavior with a plain written cue and with your mark. Alternate the conditions for a few ordinary occasions. A tiny personal comparison will not establish a universal law, but it can tell you whether the mark is useful to you in this setting.
When you return to the larger model of nested rooms, remember this book. A resemblance at one scale is a reason to ask a question at another. It is not a substitute for the answer.